The user patterns are a bit different in each case. When I'm copy-pasting from GPT, I'm forced to define the problem better and be specific on the output I'm looking for.
Compared to Cursor with full codebase context where I can get away with less, so I typically fall back to the lazy prompting patterns of "not like that, do better". Which likely eats up more time than had I prompted well to begin with.
Compared to Cursor with full codebase context where I can get away with less, so I typically fall back to the lazy prompting patterns of "not like that, do better". Which likely eats up more time than had I prompted well to begin with.